Greetings. Welcome to MARA's Third Quarter 2024 Earnings Call. At this time, all participants are in a listen-only mode. Please note, this conference is being recorded. I will now turn the conference over to your host, Robert Samuels, Vice President of Investor Relations. Thank you. You may begin.

Thank you, operator. Good afternoon, and welcome to MARA's third quarter 2024 earnings call. Thank you for joining us today. With me on today's call are our Chairman and Chief Executive Officer, Fred Thiel; and our Chief Financial Officer, Salman Khan. Certain statements made during this call may be considered forward-looking statements within the meaning of the federal securities laws. In particular, any statements about our future growth plans and performance, our liquidity position, our growth opportunities, and our future financial performance are forward-looking statements. These statements are often identified by the use of words such as anticipate, believe, estimate, intend, design, may, plan, project, would, and similar expressions or variations. Investors are cautioned not to place undue reliance on these forward-looking statements. All forward-looking statements made on today's call involve risks and uncertainties. While we may elect to update these forward-looking statements at some point in the future, we have no current intention of doing so except to the extent required by applicable law. Our actual results and outcomes may differ materially from those included in these forward-looking statements as a result of risk factors, including, but not limited to, the factors discussed under the heading Risk Factors in our most recent annual report on Form 10-K and any other periodic reports that we may file with the Securities and Exchange Commission. Finally, please note that on today's call, we will refer to certain financial measures that were not prepared in accordance with generally accepted accounting principles in the United States, including adjusted EBITDA and non-GAAP total margin. MARA believes these non-GAAP financial measures are important indicators of its operating performance because they exclude certain items that we do not believe directly reflect our core operations and may not be indicative of our recurring operations. Please refer to the earnings release for a full reconciliation of these non-GAAP financial measures to the most comparable GAAP financial measures. We hope you've had the chance to read our new Shareholder Letter and look forward to your feedback. Today's call will be a little bit different. We'll begin with some brief prepared remarks from Fred and Salman. After their comments, we are going to be conducting an analyst interview with management. We are going to be doing this on a rotating basis, and our inaugural session will be conducted by Kevin Dede, an analyst at H.C. Wainwright. Once Kevin is finished, we will go through some of the more popular questions from our retail investors. And with that out of the way, I'm going to turn the call over to Fred to kick things off. Fred?

Hello, everybody. Yesterday, we announced the addition of approximately 372 megawatts of owned and operated capacity across three sites in Ohio. These additions include the acquisition of two data centers, offering a combined 222 megawatts of interconnect-approved capacity, and the development of a third greenfield data center, which is expected to add another 150 megawatts of compute capacity. We acquired two operational data centers in Hannibal and Hopedale, Ohio with 222 megawatts of combined capacity. These sites have 122 megawatts of current capacity and are approved to expand by another 100 megawatts in 2025. Simultaneously, MARA has begun developing a 150-megawatt operation in Findlay, Ohio, which already has 30 megawatts of capacity. These three facilities have a combined interconnect-approved capacity of 372 megawatts, which MARA intends to fully energize by the end of 2025. Compute for these sites is purchased, secured, and ready for deployment. These data centers contribute to the resilience of MARA's flexible compute portfolio. This initiative increases MARA's exposure to PJM, one of the largest and most sophisticated independent system operators, or ISOs. Once energized, our data centers will be further diversified across multiple jurisdictions, ensuring that no single ISO contains more than 50% of our owned and operated capacity. We intend to further expand and diversify our portfolio of owned and operated sites, which we expect to yield significant cost savings. These growth initiatives align with our strategic goal to position MARA as one of the most cost-effective and efficient operators in the industry. And with that, I'll hand it over to Salman.

Thank you, Fred. Through our proven discipline and data center expertise, we acquired these assets at a multiple of approximately $270,000 per megawatt, based on improved capacity and after customary adjustments. This is one of the lowest disclosed multiples amongst our larger publicly traded peers, which demonstrates our unmatched ability to deliver accretive acquisitions. As the largest, best-capitalized company in our space, MARA has the privilege of reviewing every potential M&A opportunity on the market. We have been at the table for every scalable transaction announced this year, and if those opportunities traded to our peers, it was because we passed on them. Our M&A track record reflects our discipline in organic growth and our focus on maximizing shareholder value. Furthermore, these data centers will increase our total owned and operated compute capacity by over 70%. Owning the sites will provide us with greater operational control and could further reduce our operating costs at Hopedale data center, which is part of this acquisition, by up to 50%. Much of the future capacity is alongside operating generation, providing opportunities for cost-reduction, power redundancy, and development optionality. I want to remind you that in 2024 so far, MARA has secured approximately 1 gigawatt of nameplate capacity through acquisitions and greenfield site developments. As a result of strategic expansion, the company's total nameplate capacity has increased to just under 1.5 gigawatts, with approximately 65% of this capacity proudly owned and operated by MARA. With that, I will hand it back to Rob.

Thanks, Salman. Thanks, Fred. Now, I'm going to turn the call over to Kevin Dede from H.C. Wainwright to begin our interview. Kevin?

Thanks, Rob. Hi, Fred, Salman, thanks so much for having me on. I think maybe Fred, it probably makes sense to start with the deal you just closed. Why don't we talk a little bit about how you see it sort of folding into MARA's overarching thesis in driving energy costs to near zero?

Sure. So, as you look at any acquisition we do, we're focused on obviously lowering our cost of operations. These particular sites, one of which we had some capacity operating at, allow us to lower our cost of operating pretty significantly. Our goal over time, as we continue to grow our capacity, is to focus on sites where we'll continually be able to lower our operating costs. For example, if you look at the on-site generation that we recently announced through our partnership with NGON, we're able to generate electricity at a cost around $0.01 per kilowatt-hour, which is significantly lower than the typical market price of around $0.04 per kilowatt-hour paid for by miners using grid energy. As we continue to expand those opportunities, and by the way, that type of stranded gas opportunity in gigawatt-hours measured is over 1 million gigawatt-hours per year of capacity, when you look at the overall market availability of gas. We believe that you'll see miners like MARA go after these types of opportunities where we own and control our own power generation. And over time, MARA will be building more and more owned and operated generation capacity and find ways in partnering with AI companies, hyperscalers, providing services around energy management, which will allow us to subsidize our operating costs as we co-locate with them. I think over the next few quarters, you'll see us expose our plans in those areas, what we're doing, the partnerships we have, and I think it will make it very clear what our vision is as we transition from a traditional mining company that has been predominantly dependent on third-party operators to a mining company that's very focused on owning and operating its own resources and a mining company that really focuses on providing energy transition capabilities to companies way outside of mining.

Specific to the Ohio sites and their connection to PJM, what sort of flexibility do you have in your use of power and maybe an ability to resell it back to the grid?

PJM gives us the opportunity to leverage up. Power trading gives us the opportunity to leverage hedges, and gives us the opportunity to do all the things that as a large load consumer, we would want to be able to do. I think the benefits of what we've learned from how we can operate in Texas, now being able to operate with PJM being one of the largest ISOs in this space, will give us a lot of flexibility in managing power costs. As we continue to grow and expand our footprint there, it will be more and more powerful over time.

I understand you mentioned that the next few quarters will likely show more of your zero power initiative, but if we examine the progress MARA made in the third quarter, are there any highlights related to the multi-faceted strategy that MARA has implemented that could indicate a reduction in power costs?

Sure. I mean the NGON partnership is a very early first step, where you see us essentially going after micro data centers at the edge. And as I said, the opportunities there are very deep and broad across oil and gas fields domestically as well as internationally. The domestic opportunities themselves are huge. We believe we now have a core competency in being able to do this and execute it at scale. And as we continue to grow, it will become a bigger and bigger portion of our fleet and owned and operated data centers.

Okay. Let's switch topics and discuss the nearly 27,000 Bitcoin that MARA currently holds. Can you explain your HODL strategy and how it contributes to building the company's presence? What kind of financial strategies might enhance its yield? I would appreciate more insight into your HODL strategy.

Sure. So, I think one way to look at it is, we have purchased roughly 6,400 BTC at an average cost of around $60,000 per BTC. If you think about the convert transaction we did in late August, where we purchased a little over 4,000 Bitcoin, that transaction has netted a huge IRR to our shareholders over a very short period of time because of where the price of Bitcoin is today. We believed at the time, the reason we go into the market and opportunistically buy Bitcoin is that we believe there will be a price inflection point where it will be of huge value to our shareholders for us to go buy Bitcoin versus investing assets into building mining capacity, which won't generate Bitcoin for potentially 12 to 18 months. Here is a very good proof point of that strategy playing out positively for our shareholders and for Marathon. If you look over the multiple purchases of Bitcoin we've done over the years, it has netted a very healthy return for our shareholders. We believe we will continue to execute a strategy of a hybrid of buying Bitcoin when it makes sense and mining Bitcoin when it makes sense. Our average cost of the Bitcoin in our HODL today is somewhere under $50,000 when you average our cost to mine of the Bitcoin we've mined in our treasury and the Bitcoin we've acquired. If you look at that compared to today's price of Bitcoin, it's a significant appreciation in value. Today, I think we hold almost $2.5 billion in Bitcoin on the balance sheet. Now, regarding the latter part of your question, what do you do with that Bitcoin? There are ways to generate yield on that Bitcoin, which can generate income that could provide cash to buy more Bitcoin, reduce SG&A expenses, lower power costs, or make equipment purchases. This gives us financial flexibility. We can also collateralize it and borrow against it. We believe we are well-positioned for whatever the market brings. With this current jump in the price of Bitcoin, we believe the period of low-cost growth is likely coming to an end. Companies will now be forced to compete for capacity differently, and smaller miners will face more challenges, which may make consolidation a more attractive option. However, we are not in the business of paying $1.6 million a megawatt for capacity like some of our peers have done. Look at this transaction, $277,000 a megawatt, almost half of what we paid at the beginning of this year for other assets we acquired. We are going to be very smart about our strategy. We have taken advantage of machine pricing and ordered more machines than we needed for our 50 exahash goal, so we would be ready for the growth spurts next year. We believe next year will have capacity constraints for some miners, and those of us who planned in advance and have machines on order at attractive prices will be the ones dominating the marketplace in the next years leading into the next halving. We think our tech stack, our ability to build custom miners, our 2PIC immersion technology, allows us to run miners at significant levels of overclock and optimize our energy use. This will make us the predominant miner in this industry from a cost perspective.

So, Fred, apart from an opportunistic approach, how should we think about MARA's priorities and capital use between Bitcoin, on the balance sheet, acquisition of sites, deployment of more mining machines, partnerships with potential companies to reuse heat, and assorted other opportunities, not to mention software development?

Kevin, that's a big question, almost more suited for an Analyst Day type of fireside chat than we have time for today, but you can think of it this way. We have a very well-defined plan for organic expansion that combines a focus on owned and operated energy, owned and operated data centers, and collaboration with large-scale AI hyperscalers. We have a very well-defined plan around our technology stack and how we're going to deploy it. You'll see next generations of our 2PIC miners focus specifically on AI, also combined functionality for both AI and mining. You'll see different form factors of those specifically designed for edge inference locations, where we'll be able to operate anything from 4U up to 48U in a single two-phase immersion container. These systems will continue to get more efficient and leverage the increasing power and heat densities that come from the AI world. As we move up the stack towards our pool and software systems for energy balancing, I think you'll continue to see a series of announcements over the next year or two positioning MARA as one of the leading solutions providers for optimizing idle electrons, no matter where they stand.

The Ohio deal is very compelling, Fred. Were there any special negotiations behind the scenes that allowed you to elicit such favorable terms? How much more supply at that rate do you think exists in North America, given the huge demand for HPC and AI?

Well, I think it comes down to some sites being suitable for AI and HPC while others are not. It’s like that old commercial: sometimes you think like a nut, and sometimes you don’t. There are sites that are good for Bitcoin mining because they have a load that must be curtailed or lack access to high-speed internet or fiber connectivity. We believe if you look at the multiple Bitcoin mining sites built over the years and the power projects, there is a lot of low-hanging fruit in this price range. The challenge is scale. MARA is of a scale today where we're not interested in 20, 30, 40, or 50 megawatt sites. Our focus is on the 100-megawatt and larger scale, especially around what we call utility-scale mining. For on-site generation, we're very focused on micro data centers because we have the technology to deploy fault-tolerant miners and generate our own energy, which means a low overall cost to operate. As we look more towards the future, ultra-high scale or what I would call hyperscale comes from relationships with hyperscalers, where a symbiotic relationship exists between Bitcoin miners and large loads in the AI HPC space. MARA will continue scaling from power all the way through to energy management.

You mentioned 2PIC a couple of times, Fred. The letter referred to tanks you intended to shift and ship in Q4. Aside from the ones you had running in a test facility, what can you say about them and how you geared up production? I know the letter mentioned a third-party order rate already. How are you allocating resources to meet that demand while also upgrading some of your own sites?

Sure. We are fortunate to have one very large customer called MARA, who is a valued client, but they are tough negotiators. Then, we have third-party miners and data center operators interested in our technology. They have been willing to place orders with us to test the technology and assess its applicability in their systems. I think nearly 40 systems between what MARA uses and third-parties will be delivered and installed by the end of this year. We are currently operating 2PIC in production and getting good results. We're seeing overclock well over 20% in systems very efficiently. The additional advantage of 2PIC is there's a low PUE, which means the mechanical load for cooling is very low. It reduces the overall power draw of the facility, which allows even more energy for Bitcoin mining. The initial results seem attractive, and we'll continue to ramp production while iterating the design. We have refined the 2PIC solution designed specifically for Bitcoin miners, moving towards a hybrid application of miners and AI rigs with the ability for hot-swap functionality in various form factors. There is significant interest in testing this technology, and we're working with OEMs to ensure systems work appropriately. MARA has built its own custom miners for 2PIC, and we’re seeing very good results, so we'll continue to be vertically integrated and reduce dependence on offshore supply chains.

The miners you referenced here, Fred, are they the result of the work done with Auradine? Could you also comment on wafer sourcing given the controversy surrounding Bitmain and its AI partner at TSMC?

Yes, we're using Auradine miners, both their standard miners and at the silicon level, optimizing for our own applications. That being said, we've designed our 2PIC miner unit to be able to use hashboards from third-party providers as well. This is a product we’re keeping for our own use. The advantages of power densities in our systems suggest it's better that we use it internally. We have optimal flexibility; we can use Auradine chips or third-party hashboards. In terms of wafers, 3-nanometer wafers are in short supply because of the GPU market. However, we have not seen any indication from TSMC that Bitmain's supplier access to wafers has been impacted. Auradine has ample access to wafers. I don't think you'll see 2-nanometer appear on the market for about three years. I anticipate a reengineering of miner technology, where the design geometry will matter less than how the hashboard is designed. This means miners will have more options, resulting in less control from hardware vendors over time.

Well, that's encouraging to hear. The shareholder letter stated that MARA has no ambition to run AI sites. You have two new Board members with great experience in data centers and energy. How do you plan to capitalize on their experience and incorporate their insights into MARA's operations?

Kevin, I'm not going to share our five-year strategic plan with you on the phone right now. But if you look at the GPU rental service business, rental prices for H100 GPUs have dropped by 75% in six months. The technology obsolescence in Bitcoin mining is aggressive, but it's even more so for GPUs. Anyone transitioning from Bitcoin mining to hyperscaler or large-scale AI hosting will encounter an abrupt surprise in two or three years as hyperscalers roll out their data centers built to operate at lower costs. Hyperscalers can choose the first hardware, power, and customers. Companies tying up agreements today, as long as those can be enforced, are in a fortunate position. We consider ourselves a technology company focused on generating high-value solutions for our shareholders rather than chasing hype trends. Our team and Board add talent and knowledge. You can start to see the direction of the technologies that MARA will bring to market, setting us apart as more than just a Bitcoin miner.

Fair enough, Fred. Thanks. We should discuss the events last week and the reflection in Bitcoin price. I’m curious about your perspective, considering MARA has been involved in the process. How do you see the anticipated changes in the industry? What forward-looking aspects do you see for the U.S. government moving to a more crypto-friendly stance? How can MARA capitalize on this?

Absolutely. One great advantage of the United States form of government is having three branches. My focus is on the legislative and executive branch. Following the election, we expect many incoming legislators to be pro-crypto. The crypto industry spent over $100 million in targeted campaigns. MARA, along with two other peers, co-founded the Bitcoin Voter Project to promote pro-crypto candidates. A major race saw crypto activists successfully unseat a prominent anti-crypto figure. We anticipate that the legislative branch will adopt a more open perspective on crypto, enabling legislation for market structure and stablecoins, as well as regulatory frameworks for Bitcoin and crypto that allow banks to handle these assets. In terms of the executive branch, indications suggest that if the Vice President intends to continue exiting similar policies to President Biden, it signals the war on crypto will persist. In contrast, President Trump has been vocal about his support for Bitcoin. There are proposals to establish a strategic reserve while recognizing Bitcoin's strategic importance. I will advocate for initiatives that enable the free use of Bitcoin, equal access to power for miners, and promote Bitcoin as an investment asset.

Lastly, Salman, looking back at MARA's overall financial performance, what highlights can you share? Where do you think improvement is needed moving forward, and what can we expect in the future?

Thank you, Kevin, for the question. As we've reported today, our energized hash rate has increased 93% to 36.9 exahash per second in Q3 of this year compared to Q3 last year. Our production is public information after quarter-end as well. So, we've exceeded 40 exahash already, and we are driving towards the 50 exahash target set for ourselves earlier this year. Regarding total block space, we achieved a significant highlight: despite halving, we achieved a 32% increase compared to Q2 of 2024. This increase directly impacts our revenues, which saw a 35% increase compared to the same quarter last year. Influencing factors include Bitcoin price fluctuations, halving, and the global hash rate. Despite these challenges, our revenues increased significantly compared to last year. On the cost side, the story is similar. Our cost of revenue per petahash per day perspective has improved throughout this year. Acquisitions and organic growth contributed to driving costs down. The cash cost of revenue, a non-GAAP measure disclosed in our financials, has decreased by 18% from $45.2 in Q3 2023 to $37.1 per petahash per day in Q3 2024, reflecting our near-zero energy cost objective. As a reminder, we were an asset-light company just a year ago. With the gigawatt addition in capacity, once fully energized, we anticipate these acquisitions will provide benefits over time. Sequentially, we have continued to reduce costs quarter-over-quarter. In the balance sheet, as Fred mentioned, we now hold over 27,000 Bitcoin post-quarter-end, resulting in a considerable rise in value. Regarding the convert transaction, we have multiple sources of capital available, which is encouraging. Though it's premature in the industry, we expect more sources to open up as we evolve. Our decision to buy as well as produce Bitcoin has been highly accretive for us, demonstrated by the 6,400 Bitcoin purchased this year at an average price of $60,000 per Bitcoin, resulting in approximately 50% increases in our BTC yield per share. This acquisition strategy and our strong balance sheet continues to position us for long-term value creation.

Thanks, Salman. I imagine we'll see that decline again through the fourth quarter as the Ohio sites come online. Thank you very much for entertaining my questions, gentlemen. I really appreciate it. And Fred, thanks for digesting the questions and offering a shorter answer than you may have wanted to otherwise. I appreciate that. I'll turn the call back to Rob. Thanks again for having me on.

Thank you.

Thanks, Kevin. We really appreciate all the questions. As usual, we received some questions from our retail investors. Most of them hit on some similar questions you were asking, Kevin. There is one thing that people are asking that I would like to address to Fred: how involved do you expect MARA to be in shaping U.S. Bitcoin and Bitcoin mining/energy policy with the incoming administration?

I believe members of the incoming administration will engage with Bitcoin miners regarding various policies, and we'll continue to advise and consult and provide input when asked. We're committed to taking an active role. We're focused on collaborating with all branches of government, whether it be the executive branch, the legislative branch, or departments like FERC and DOE. We aim to educate them on how Bitcoin mining can benefit the energy grid and transition to a more electrified nation. This country will need to invest trillions to upgrade its grid as it continues to deploy renewable energy. Without a broad acceptance of further nuclear energy in this country, all renewable energy will be intermittent, necessitating dispatchable loads to balance the load. Bitcoin mining is in a unique position to fulfill this need, and we will work with the government to ensure this consideration is recognized as they plan and build out the grid.

Thanks, Fred, and thank you everyone for your time today. If you have questions that were not answered during today's call, please feel free to contact our Investor Relations team at [email protected]. Thanks very much, and enjoy the rest of the day.

Thank you. All parties may disconnect.